DNA methylation estimates based on WGBS, RRBS and MREBS data in different chromatin states.

A. Heat map and functional annotation for an 18-level chromatin state model at 200 bp resolution built using peak calls made using ChIP-seq data sets for 10 histone modifications, as well as an input library, for each of the three cell types described in Fig 1A: i) MEFs, ii) EARLY intermediates, and iii) ESCs, as well as a late reprogramming intermediate (LATE), partially induced pluripotent cells, or pre-iPSCs, not otherwise used in the study. Candidate functional annotations were assigned to each of the 18 chromatins states based on the prevalence and combination of histone mark peaks, which could in turn be classified into the seven categories indicated in the left-hand column. The probability of a window in each state to be contain a peak for a given histone modification is given as a percentage in each cell, and visually indicated by the intensity of color in the heat map. The proportion of the concatenated genome (MEFs + EARLY + LATE + ESCs) found in each of the 18 chromatin states is given in the third column. B. Violin plots of the distributions of the DNA methylation estimates in each of the 18 chromatin states (described in A), as well as genome-wide (blue), for EARLY intermediates using i) whole-genome bisulfite sequencing (WGBS), reduced representation bisulfite sequencing (RRBS), and iii) methylation-sensitive restriction enzyme bisulfite sequencing (MREBS). The mean DNA methylation estimates within 200 bp windows corresponding to those used for the chromatin state model were used, only considering those windows containing at least one CpG with 5X coverage, in an effort to ensure high-confidence estimates. White circles represent median values. C. As in (B), but for ESCs.